fix[gpu]: decode extension arrays on the GPU

[0ax1]

Fixes #8143

[codspeed-hq]

Merging this PR will not alter performance

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚡ 6 improved benchmarks
❌ 6 regressed benchmarks
✅ 1520 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	chunked_bool_canonical_into[(1000, 10)]	20.2 µs	35.2 µs	-42.51%
❌	Simulation	chunked_varbinview_into_canonical[(1000, 10)]	177.2 µs	213.3 µs	-16.93%
❌	Simulation	varbinview_large	113.1 µs	131.7 µs	-14.14%
❌	Simulation	decompress_rd[f64, (100000, 0.0)]	845.2 µs	980 µs	-13.76%
❌	Simulation	bitwise_not_vortex_buffer_mut[128]	215.3 ns	244.4 ns	-11.93%
❌	Simulation	chunked_varbinview_canonical_into[(100, 100)]	274.4 µs	309.3 µs	-11.27%
⚡	Simulation	decompress_rd[f64, (10000, 0.0)]	138 µs	110.9 µs	+24.41%
⚡	Simulation	decompress_rd[f64, (10000, 0.1)]	137.8 µs	110.9 µs	+24.24%
⚡	Simulation	decompress_rd[f64, (10000, 0.01)]	137.4 µs	110.6 µs	+24.23%
⚡	Simulation	decompress_rd[f32, (10000, 0.1)]	89.3 µs	80.2 µs	+11.4%
⚡	Simulation	decompress_rd[f32, (10000, 0.0)]	89.6 µs	80.8 µs	+10.93%
⚡	Simulation	decompress_rd[f32, (10000, 0.01)]	89.3 µs	80.7 µs	+10.66%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing ad/gpu-decode-ext-storage (2d6871c) with develop (3d7bbfb)}